What is “IP warming” and how does it affect reputation?

Still have a question, spotted an error, or have a better explanation or a source we should cite?

You've just switched to a dedicated IP, or you're migrating to a new ESP. Your first instinct is to send your whole list right away. Don't. That's exactly how you end up in the spam folder for the next two months.

IP warming is the practice of gradually increasing your sending volume on a new IP address so mailbox providers can build a picture of who you are before they trust you with high volumes. A brand-new IP has zero history. Gmail, Outlook, and Yahoo Mail treat unknown IPs the way a bank treats a brand-new customer: cautiously, until there's enough data to go on.

The full process typically takes 4 to 8 weeks, depending on your target volume. If you're aiming to send a million emails a month, expect to be closer to 8 weeks. If you're a smaller sender in the tens of thousands, you can warm up faster.

What a warming schedule actually looks like

The principle is simple: start small, send to your most engaged subscribers first, and double (roughly) every few days if things go well. Here's a rough framework:

  • Days 1-3: 200-500 emails per day. Your absolute best subscribers. People who open everything.
  • Days 4-7: 1,000-2,000 per day. Still high-engagement contacts only.
  • Week 2: 5,000-10,000 per day. Start broadening to anyone who's opened in the last 90 days.
  • Week 3-4: 20,000-50,000 per day. You can start including less-recent openers, but watch your metrics closely.
  • Week 5-8: Scale toward full volume, adjusting based on what the data tells you.

These numbers aren't rigid. They're a starting point. Your actual ramp depends on how your sending reputation develops. If you see strong open rates and very few complaints, you can accelerate. If you're seeing throttling signals or spam folder placement, slow down.

Why reputation mechanics matter here

During warming, every delivery is a data point. Mailbox providers are watching your open rate, your complaint rate, your bounce rate, and whether recipients are marking your mail as spam or deleting it unread. High engagement early on tells them this IP sends mail people actually want. Low engagement or complaints tells them the opposite.

And this is why you always start with your most engaged subscribers. They give you the best possible signal. Sending to a cold or unvalidated list during warmup is the fastest way to destroy a reputation before it even forms. If your list hasn't been cleaned recently, that's worth sorting out before you start. (We clean lists at RME Clean if you need a hand ;)

Signs you're warming too fast

  • Emails going to spam folders instead of the inbox
  • Unusually low open rates compared to your historical average
  • SMTP deferrals or rate-limit errors in your sending logs
  • Complaint rates above 0.08% in any single send
  • Sudden increase in soft bounces from major providers

If any of those show up, pause, drop back to a lower volume for a few days, and let the reputation stabilise before pushing up again.

Authentication is not optional during warming

Your IP reputation doesn't build in isolation. SPF, DKIM, and DMARC need to be fully configured before day one of warming. Providers use authentication to confirm that the IP and domain match up. Warming without proper authentication is like trying to open a bank account without ID. You won't get far.

Not sure if your authentication is set up correctly? Our free SPF checker takes 30 seconds and will tell you if something's off before you start sending.

Contributors

Who worked on this answer

Every name links to their profile. Every company links to their site. Real people, real accountability.

Ask an AI · tailored to your setup

Get your custom warming schedule

I'm warming up a new sending IP for your sending domain using your ESP name. My target sending volume is emails per month. My list has total list size contacts and my recent average open rate is open rate %. Based on this, can you give me a week-by-week warming schedule with daily send caps? Also flag the key metrics I should watch each week and the thresholds that should make me slow down.

Edit the yellow boxes, then send to the AI of your choice.